FCDdetection

The following scripts accompany the publication: "Novel surface features for automated detection of focal cortical dysplasias in paediatric epilepsy." http://dx.doi.org/10.1016/j.nicl.2016.12.030 Briefly, the scripts calculate surface-based structural MRI features from cortical reconstructions and use these data to train a supervised neural network classifier to identify lesion-like vertices. In the sample used, the leave-one-out classifier was able to correctly identify FCDs in 73% of patients.

Please send any queries to kw350@cam.ac.uk or sophie.adler.13@ucl.ac.uk.

The original scans could not be shared publicly, but the matrix of subjects' morphological data, along with lesion/non-lesion labelling of each vertex, can be freely downloaded from: https://doi.org/10.17863/CAM.6923

The scripts are numbered 1-10 Pre-script steps:

You need to have FreeSurfer cortical reconstructions of all your participants (https://surfer.nmr.mgh.harvard.edu/). It is important that these are checked and edits are done to correct the surfaces.

it is important to check that the FLAIR scan is correctly coregistered to the T1 scan and therefore to the surfaces.
With volumetric FLAIR, the recon-all process included the FLAIR scan. If volumetric FLAIR is unavailable, supplementary script 1 will coregister the FLAIR scan after the recon-all step (Supplementary_script_1). Further analyses need to be made to assess whether non-volumetric FLAIR is sufficient.

Create manual lesion labels of the FCDs.

this can be done in FreeSurfer
after creating the labels, they need to be converted to .mgh files for compatibility with the rest of the scripts (see Supplementary_script_2).

Script 1: This script does the following

Sample FLAIR at 25%, 50%, 75% of the cortical thickness, at the grey-white matter boundary, and 0.5mm and 1mm subcortically
Smooth the following features using a 10mm gaussian kernel:

cortical thickness
FLAIR at 25%, 50%, 75% of the cortical thickness, at the grey-white matter boundary, and 0.5mm and 1mm subcortically
grey-white matter intensity contrast

Calculate curvature for use in script 2 - calculation of local cortical deformation
Convert curvature and sulcal depth and lgi to .mgh file type

Script 2: This script calculates local cortical deformation

Script 3: This script calculates the Doughnut method

Script 4: Smoothing of local cortical deformation and doughnut metrics

Script 5: Intra-subject normalisation of features

Script 6: Move features to template space (this involves flipping the right hemisphere features so that everything is moved to the left hemisphere)

Script 7: Inter-subject normalisation of features for the classifier

Script 8: Neural Network classifier (including principal component analysis for determining number of nodes in classifier)

Script 9: Clustering of classifier output

Script 10: Ranking of top 5 clusters

FCDdetection

The following scripts accompany the publication: "Novel surface features for automated detection of focal cortical dysplasias in paediatric epilepsy." http://dx.doi.org/10.1016/j.nicl.2016.12.030 Briefly, the scripts calculate surface-based structural MRI features from cortical reconstructions and use these data to train a supervised neural network classifier to identify lesion-like vertices. In the sample used, the leave-one-out classifier was able to correctly identify FCDs in 73% of patients.

Please send any queries to kw350@cam.ac.uk or sophie.adler.13@ucl.ac.uk.

The original scans could not be shared publicly, but the matrix of subjects' morphological data, along with lesion/non-lesion labelling of each vertex, can be freely downloaded from: https://doi.org/10.17863/CAM.6923

The scripts are numbered 1-10 Pre-script steps:

You need to have FreeSurfer cortical reconstructions of all your participants (https://surfer.nmr.mgh.harvard.edu/). It is important that these are checked and edits are done to correct the surfaces.

it is important to check that the FLAIR scan is correctly coregistered to the T1 scan and therefore to the surfaces.
With volumetric FLAIR, the recon-all process included the FLAIR scan. If volumetric FLAIR is unavailable, supplementary script 1 will coregister the FLAIR scan after the recon-all step (Supplementary_script_1). Further analyses need to be made to assess whether non-volumetric FLAIR is sufficient.

Create manual lesion labels of the FCDs.

this can be done in FreeSurfer
after creating the labels, they need to be converted to .mgh files for compatibility with the rest of the scripts (see Supplementary_script_2).

Script 1: This script does the following

Sample FLAIR at 25%, 50%, 75% of the cortical thickness, at the grey-white matter boundary, and 0.5mm and 1mm subcortically
Smooth the following features using a 10mm gaussian kernel:

cortical thickness
FLAIR at 25%, 50%, 75% of the cortical thickness, at the grey-white matter boundary, and 0.5mm and 1mm subcortically
grey-white matter intensity contrast

Calculate curvature for use in script 2 - calculation of local cortical deformation
Convert curvature and sulcal depth and lgi to .mgh file type

Script 2: This script calculates local cortical deformation

Script 3: This script calculates the Doughnut method

Script 4: Smoothing of local cortical deformation and doughnut metrics

Script 5: Intra-subject normalisation of features

Script 6: Move features to template space (this involves flipping the right hemisphere features so that everything is moved to the left hemisphere)

Script 7: Inter-subject normalisation of features for the classifier

Script 8: Neural Network classifier (including principal component analysis for determining number of nodes in classifier)

Script 9: Clustering of classifier output

Script 10: Ranking of top 5 clusters

FCDdetection

The following scripts accompany the publication: "Novel surface features for automated detection of focal cortical dysplasias in paediatric epilepsy." http://dx.doi.org/10.1016/j.nicl.2016.12.030 Briefly, the scripts calculate surface-based structural MRI features from cortical reconstructions and use these data to train a supervised neural network classifier to identify lesion-like vertices. In the sample used, the leave-one-out classifier was able to correctly identify FCDs in 73% of patients.

Please send any queries to kw350@cam.ac.uk or sophie.adler.13@ucl.ac.uk.

The original scans could not be shared publicly, but the matrix of subjects' morphological data, along with lesion/non-lesion labelling of each vertex, can be freely downloaded from: https://doi.org/10.17863/CAM.6923

The scripts are numbered 1-10 Pre-script steps:

You need to have FreeSurfer cortical reconstructions of all your participants (https://surfer.nmr.mgh.harvard.edu/). It is important that these are checked and edits are done to correct the surfaces.

it is important to check that the FLAIR scan is correctly coregistered to the T1 scan and therefore to the surfaces.
With volumetric FLAIR, the recon-all process included the FLAIR scan. If volumetric FLAIR is unavailable, supplementary script 1 will coregister the FLAIR scan after the recon-all step (Supplementary_script_1). Further analyses need to be made to assess whether non-volumetric FLAIR is sufficient.

Create manual lesion labels of the FCDs.

this can be done in FreeSurfer
after creating the labels, they need to be converted to .mgh files for compatibility with the rest of the scripts (see Supplementary_script_2).

Script 1: This script does the following

Sample FLAIR at 25%, 50%, 75% of the cortical thickness, at the grey-white matter boundary, and 0.5mm and 1mm subcortically
Smooth the following features using a 10mm gaussian kernel:

cortical thickness
FLAIR at 25%, 50%, 75% of the cortical thickness, at the grey-white matter boundary, and 0.5mm and 1mm subcortically
grey-white matter intensity contrast

Calculate curvature for use in script 2 - calculation of local cortical deformation
Convert curvature and sulcal depth and lgi to .mgh file type

Script 2: This script calculates local cortical deformation

Script 3: This script calculates the Doughnut method

Script 4: Smoothing of local cortical deformation and doughnut metrics

Script 5: Intra-subject normalisation of features

Script 6: Move features to template space (this involves flipping the right hemisphere features so that everything is moved to the left hemisphere)

Script 7: Inter-subject normalisation of features for the classifier

Script 8: Neural Network classifier (including principal component analysis for determining number of nodes in classifier)

Script 9: Clustering of classifier output

Script 10: Ranking of top 5 clusters

FCDdetection

The following scripts accompany the publication: "Novel surface features for automated detection of focal cortical dysplasias in paediatric epilepsy." http://dx.doi.org/10.1016/j.nicl.2016.12.030 Briefly, the scripts calculate surface-based structural MRI features from cortical reconstructions and use these data to train a supervised neural network classifier to identify lesion-like vertices. In the sample used, the leave-one-out classifier was able to correctly identify FCDs in 73% of patients.

Please send any queries to kw350@cam.ac.uk or sophie.adler.13@ucl.ac.uk.

The original scans could not be shared publicly, but the matrix of subjects' morphological data, along with lesion/non-lesion labelling of each vertex, can be freely downloaded from: https://doi.org/10.17863/CAM.6923

The scripts are numbered 1-10 Pre-script steps:

You need to have FreeSurfer cortical reconstructions of all your participants (https://surfer.nmr.mgh.harvard.edu/). It is important that these are checked and edits are done to correct the surfaces.

it is important to check that the FLAIR scan is correctly coregistered to the T1 scan and therefore to the surfaces.
With volumetric FLAIR, the recon-all process included the FLAIR scan. If volumetric FLAIR is unavailable, supplementary script 1 will coregister the FLAIR scan after the recon-all step (Supplementary_script_1). Further analyses need to be made to assess whether non-volumetric FLAIR is sufficient.

Create manual lesion labels of the FCDs.

this can be done in FreeSurfer
after creating the labels, they need to be converted to .mgh files for compatibility with the rest of the scripts (see Supplementary_script_2).

Script 1: This script does the following

Sample FLAIR at 25%, 50%, 75% of the cortical thickness, at the grey-white matter boundary, and 0.5mm and 1mm subcortically
Smooth the following features using a 10mm gaussian kernel:

cortical thickness
FLAIR at 25%, 50%, 75% of the cortical thickness, at the grey-white matter boundary, and 0.5mm and 1mm subcortically
grey-white matter intensity contrast

Calculate curvature for use in script 2 - calculation of local cortical deformation
Convert curvature and sulcal depth and lgi to .mgh file type

Script 2: This script calculates local cortical deformation

Script 3: This script calculates the Doughnut method

Script 4: Smoothing of local cortical deformation and doughnut metrics

Script 5: Intra-subject normalisation of features

Script 6: Move features to template space (this involves flipping the right hemisphere features so that everything is moved to the left hemisphere)

Script 7: Inter-subject normalisation of features for the classifier

Script 8: Neural Network classifier (including principal component analysis for determining number of nodes in classifier)

Script 9: Clustering of classifier output

Script 10: Ranking of top 5 clusters

FCDdetection

The following scripts accompany the publication: "Novel surface features for automated detection of focal cortical dysplasias in paediatric epilepsy." http://dx.doi.org/10.1016/j.nicl.2016.12.030 Briefly, the scripts calculate surface-based structural MRI features from cortical reconstructions and use these data to train a supervised neural network classifier to identify lesion-like vertices. In the sample used, the leave-one-out classifier was able to correctly identify FCDs in 73% of patients.

Please send any queries to kw350@cam.ac.uk or sophie.adler.13@ucl.ac.uk.

The original scans could not be shared publicly, but the matrix of subjects' morphological data, along with lesion/non-lesion labelling of each vertex, can be freely downloaded from: https://doi.org/10.17863/CAM.6923

The scripts are numbered 1-10 Pre-script steps:

You need to have FreeSurfer cortical reconstructions of all your participants (https://surfer.nmr.mgh.harvard.edu/). It is important that these are checked and edits are done to correct the surfaces.

it is important to check that the FLAIR scan is correctly coregistered to the T1 scan and therefore to the surfaces.
With volumetric FLAIR, the recon-all process included the FLAIR scan. If volumetric FLAIR is unavailable, supplementary script 1 will coregister the FLAIR scan after the recon-all step (Supplementary_script_1). Further analyses need to be made to assess whether non-volumetric FLAIR is sufficient.

Create manual lesion labels of the FCDs.

this can be done in FreeSurfer
after creating the labels, they need to be converted to .mgh files for compatibility with the rest of the scripts (see Supplementary_script_2).

Script 1: This script does the following

Sample FLAIR at 25%, 50%, 75% of the cortical thickness, at the grey-white matter boundary, and 0.5mm and 1mm subcortically
Smooth the following features using a 10mm gaussian kernel:

cortical thickness
FLAIR at 25%, 50%, 75% of the cortical thickness, at the grey-white matter boundary, and 0.5mm and 1mm subcortically
grey-white matter intensity contrast

Calculate curvature for use in script 2 - calculation of local cortical deformation
Convert curvature and sulcal depth and lgi to .mgh file type

Script 2: This script calculates local cortical deformation

Script 3: This script calculates the Doughnut method

Script 4: Smoothing of local cortical deformation and doughnut metrics

Script 5: Intra-subject normalisation of features

Script 6: Move features to template space (this involves flipping the right hemisphere features so that everything is moved to the left hemisphere)

Script 7: Inter-subject normalisation of features for the classifier

Script 8: Neural Network classifier (including principal component analysis for determining number of nodes in classifier)

Script 9: Clustering of classifier output

Script 10: Ranking of top 5 clusters

FCDdetection

The following scripts accompany the publication: "Novel surface features for automated detection of focal cortical dysplasias in paediatric epilepsy." http://dx.doi.org/10.1016/j.nicl.2016.12.030 Briefly, the scripts calculate surface-based structural MRI features from cortical reconstructions and use these data to train a supervised neural network classifier to identify lesion-like vertices. In the sample used, the leave-one-out classifier was able to correctly identify FCDs in 73% of patients.

Please send any queries to kw350@cam.ac.uk or sophie.adler.13@ucl.ac.uk.

The original scans could not be shared publicly, but the matrix of subjects' morphological data, along with lesion/non-lesion labelling of each vertex, can be freely downloaded from: https://doi.org/10.17863/CAM.6923

The scripts are numbered 1-10 Pre-script steps:

You need to have FreeSurfer cortical reconstructions of all your participants (https://surfer.nmr.mgh.harvard.edu/). It is important that these are checked and edits are done to correct the surfaces.

it is important to check that the FLAIR scan is correctly coregistered to the T1 scan and therefore to the surfaces.
With volumetric FLAIR, the recon-all process included the FLAIR scan. If volumetric FLAIR is unavailable, supplementary script 1 will coregister the FLAIR scan after the recon-all step (Supplementary_script_1). Further analyses need to be made to assess whether non-volumetric FLAIR is sufficient.

Create manual lesion labels of the FCDs.

this can be done in FreeSurfer
after creating the labels, they need to be converted to .mgh files for compatibility with the rest of the scripts (see Supplementary_script_2).

Script 1: This script does the following

Sample FLAIR at 25%, 50%, 75% of the cortical thickness, at the grey-white matter boundary, and 0.5mm and 1mm subcortically
Smooth the following features using a 10mm gaussian kernel:

cortical thickness
FLAIR at 25%, 50%, 75% of the cortical thickness, at the grey-white matter boundary, and 0.5mm and 1mm subcortically
grey-white matter intensity contrast

Calculate curvature for use in script 2 - calculation of local cortical deformation
Convert curvature and sulcal depth and lgi to .mgh file type

Script 2: This script calculates local cortical deformation

Script 3: This script calculates the Doughnut method

Script 4: Smoothing of local cortical deformation and doughnut metrics

Script 5: Intra-subject normalisation of features

Script 6: Move features to template space (this involves flipping the right hemisphere features so that everything is moved to the left hemisphere)

Script 7: Inter-subject normalisation of features for the classifier

Script 8: Neural Network classifier (including principal component analysis for determining number of nodes in classifier)

Script 9: Clustering of classifier output

Script 10: Ranking of top 5 clusters

FCDdetection

The following scripts accompany the publication: "Novel surface features for automated detection of focal cortical dysplasias in paediatric epilepsy." http://dx.doi.org/10.1016/j.nicl.2016.12.030 Briefly, the scripts calculate surface-based structural MRI features from cortical reconstructions and use these data to train a supervised neural network classifier to identify lesion-like vertices. In the sample used, the leave-one-out classifier was able to correctly identify FCDs in 73% of patients.

Please send any queries to kw350@cam.ac.uk or sophie.adler.13@ucl.ac.uk.

The original scans could not be shared publicly, but the matrix of subjects' morphological data, along with lesion/non-lesion labelling of each vertex, can be freely downloaded from: https://doi.org/10.17863/CAM.6923

The scripts are numbered 1-10 Pre-script steps:

You need to have FreeSurfer cortical reconstructions of all your participants (https://surfer.nmr.mgh.harvard.edu/). It is important that these are checked and edits are done to correct the surfaces.

it is important to check that the FLAIR scan is correctly coregistered to the T1 scan and therefore to the surfaces.
With volumetric FLAIR, the recon-all process included the FLAIR scan. If volumetric FLAIR is unavailable, supplementary script 1 will coregister the FLAIR scan after the recon-all step (Supplementary_script_1). Further analyses need to be made to assess whether non-volumetric FLAIR is sufficient.

Create manual lesion labels of the FCDs.

this can be done in FreeSurfer
after creating the labels, they need to be converted to .mgh files for compatibility with the rest of the scripts (see Supplementary_script_2).

Script 1: This script does the following

Sample FLAIR at 25%, 50%, 75% of the cortical thickness, at the grey-white matter boundary, and 0.5mm and 1mm subcortically
Smooth the following features using a 10mm gaussian kernel:

cortical thickness
FLAIR at 25%, 50%, 75% of the cortical thickness, at the grey-white matter boundary, and 0.5mm and 1mm subcortically
grey-white matter intensity contrast

Calculate curvature for use in script 2 - calculation of local cortical deformation
Convert curvature and sulcal depth and lgi to .mgh file type

Script 2: This script calculates local cortical deformation

Script 3: This script calculates the Doughnut method

Script 4: Smoothing of local cortical deformation and doughnut metrics

Script 5: Intra-subject normalisation of features

Script 6: Move features to template space (this involves flipping the right hemisphere features so that everything is moved to the left hemisphere)

Script 7: Inter-subject normalisation of features for the classifier

Script 8: Neural Network classifier (including principal component analysis for determining number of nodes in classifier)

Script 9: Clustering of classifier output

Script 10: Ranking of top 5 clusters

FCDdetection

The following scripts accompany the publication: "Novel surface features for automated detection of focal cortical dysplasias in paediatric epilepsy." http://dx.doi.org/10.1016/j.nicl.2016.12.030 Briefly, the scripts calculate surface-based structural MRI features from cortical reconstructions and use these data to train a supervised neural network classifier to identify lesion-like vertices. In the sample used, the leave-one-out classifier was able to correctly identify FCDs in 73% of patients.

Please send any queries to kw350@cam.ac.uk or sophie.adler.13@ucl.ac.uk.

The original scans could not be shared publicly, but the matrix of subjects' morphological data, along with lesion/non-lesion labelling of each vertex, can be freely downloaded from: https://doi.org/10.17863/CAM.6923

The scripts are numbered 1-10 Pre-script steps:

You need to have FreeSurfer cortical reconstructions of all your participants (https://surfer.nmr.mgh.harvard.edu/). It is important that these are checked and edits are done to correct the surfaces.

it is important to check that the FLAIR scan is correctly coregistered to the T1 scan and therefore to the surfaces.
With volumetric FLAIR, the recon-all process included the FLAIR scan. If volumetric FLAIR is unavailable, supplementary script 1 will coregister the FLAIR scan after the recon-all step (Supplementary_script_1). Further analyses need to be made to assess whether non-volumetric FLAIR is sufficient.

Create manual lesion labels of the FCDs.

this can be done in FreeSurfer
after creating the labels, they need to be converted to .mgh files for compatibility with the rest of the scripts (see Supplementary_script_2).

Script 1: This script does the following

Sample FLAIR at 25%, 50%, 75% of the cortical thickness, at the grey-white matter boundary, and 0.5mm and 1mm subcortically
Smooth the following features using a 10mm gaussian kernel:

cortical thickness
FLAIR at 25%, 50%, 75% of the cortical thickness, at the grey-white matter boundary, and 0.5mm and 1mm subcortically
grey-white matter intensity contrast

Calculate curvature for use in script 2 - calculation of local cortical deformation
Convert curvature and sulcal depth and lgi to .mgh file type

Script 2: This script calculates local cortical deformation

Script 3: This script calculates the Doughnut method

Script 4: Smoothing of local cortical deformation and doughnut metrics

Script 5: Intra-subject normalisation of features

Script 6: Move features to template space (this involves flipping the right hemisphere features so that everything is moved to the left hemisphere)

Script 7: Inter-subject normalisation of features for the classifier

Script 8: Neural Network classifier (including principal component analysis for determining number of nodes in classifier)

Script 9: Clustering of classifier output

Script 10: Ranking of top 5 clusters

Name		Name	Last commit message	Last commit date
LICENCE.txt		LICENCE.txt
README.md		README.md
Script10_Rank_top_cluster.m		Script10_Rank_top_cluster.m
Script10_Rank_top_cluster_5.m		Script10_Rank_top_cluster_5.m
Script1_sample_FLAIR_smooth_features.sh		Script1_sample_FLAIR_smooth_features.sh
Script2_local_cortical_deformation.m		Script2_local_cortical_deformation.m
Script3_doughnut_method.m		Script3_doughnut_method.m
Script4_smooth_Doughnut_LCD.sh		Script4_smooth_Doughnut_LCD.sh
Script6_move_to_xhemi_flip.sh		Script6_move_to_xhemi_flip.sh
Script7_Z_Score_features_by_controls.m		Script7_Z_Score_features_by_controls.m
Script8_NeuralNetwork.m		Script8_NeuralNetwork.m
Script9_Cluster_NN_output.m		Script9_Cluster_NN_output.m
Scripts_5_intrasubject_normalisation.m		Scripts_5_intrasubject_normalisation.m
Supplementary_script_1_register_FLAIR_to_T1.sh		Supplementary_script_1_register_FLAIR_to_T1.sh
Supplementary_script_2_convert_label_to_mgh.m		Supplementary_script_2_convert_label_to_mgh.m

Name		Name	Last commit message	Last commit date
LICENCE.txt		LICENCE.txt
README.md		README.md
Script10_Rank_top_cluster.m		Script10_Rank_top_cluster.m
Script10_Rank_top_cluster_5.m		Script10_Rank_top_cluster_5.m
Script1_sample_FLAIR_smooth_features.sh		Script1_sample_FLAIR_smooth_features.sh
Script2_local_cortical_deformation.m		Script2_local_cortical_deformation.m
Script3_doughnut_method.m		Script3_doughnut_method.m
Script4_smooth_Doughnut_LCD.sh		Script4_smooth_Doughnut_LCD.sh
Script6_move_to_xhemi_flip.sh		Script6_move_to_xhemi_flip.sh
Script7_Z_Score_features_by_controls.m		Script7_Z_Score_features_by_controls.m
Script8_NeuralNetwork.m		Script8_NeuralNetwork.m
Script9_Cluster_NN_output.m		Script9_Cluster_NN_output.m
Scripts_5_intrasubject_normalisation.m		Scripts_5_intrasubject_normalisation.m
Supplementary_script_1_register_FLAIR_to_T1.sh		Supplementary_script_1_register_FLAIR_to_T1.sh
Supplementary_script_2_convert_label_to_mgh.m		Supplementary_script_2_convert_label_to_mgh.m

Name		Name	Last commit message	Last commit date
LICENCE.txt		LICENCE.txt
README.md		README.md
Script10_Rank_top_cluster.m		Script10_Rank_top_cluster.m
Script10_Rank_top_cluster_5.m		Script10_Rank_top_cluster_5.m
Script1_sample_FLAIR_smooth_features.sh		Script1_sample_FLAIR_smooth_features.sh
Script2_local_cortical_deformation.m		Script2_local_cortical_deformation.m
Script3_doughnut_method.m		Script3_doughnut_method.m
Script4_smooth_Doughnut_LCD.sh		Script4_smooth_Doughnut_LCD.sh
Script6_move_to_xhemi_flip.sh		Script6_move_to_xhemi_flip.sh
Script7_Z_Score_features_by_controls.m		Script7_Z_Score_features_by_controls.m
Script8_NeuralNetwork.m		Script8_NeuralNetwork.m
Script9_Cluster_NN_output.m		Script9_Cluster_NN_output.m
Scripts_5_intrasubject_normalisation.m		Scripts_5_intrasubject_normalisation.m
Supplementary_script_1_register_FLAIR_to_T1.sh		Supplementary_script_1_register_FLAIR_to_T1.sh
Supplementary_script_2_convert_label_to_mgh.m		Supplementary_script_2_convert_label_to_mgh.m

Name		Name	Last commit message	Last commit date
LICENCE.txt		LICENCE.txt
README.md		README.md
Script10_Rank_top_cluster.m		Script10_Rank_top_cluster.m
Script10_Rank_top_cluster_5.m		Script10_Rank_top_cluster_5.m
Script1_sample_FLAIR_smooth_features.sh		Script1_sample_FLAIR_smooth_features.sh
Script2_local_cortical_deformation.m		Script2_local_cortical_deformation.m
Script3_doughnut_method.m		Script3_doughnut_method.m
Script4_smooth_Doughnut_LCD.sh		Script4_smooth_Doughnut_LCD.sh
Script6_move_to_xhemi_flip.sh		Script6_move_to_xhemi_flip.sh
Script7_Z_Score_features_by_controls.m		Script7_Z_Score_features_by_controls.m
Script8_NeuralNetwork.m		Script8_NeuralNetwork.m
Script9_Cluster_NN_output.m		Script9_Cluster_NN_output.m
Scripts_5_intrasubject_normalisation.m		Scripts_5_intrasubject_normalisation.m
Supplementary_script_1_register_FLAIR_to_T1.sh		Supplementary_script_1_register_FLAIR_to_T1.sh
Supplementary_script_2_convert_label_to_mgh.m		Supplementary_script_2_convert_label_to_mgh.m

Name		Name	Last commit message	Last commit date
LICENCE.txt		LICENCE.txt
README.md		README.md
Script10_Rank_top_cluster.m		Script10_Rank_top_cluster.m
Script10_Rank_top_cluster_5.m		Script10_Rank_top_cluster_5.m
Script1_sample_FLAIR_smooth_features.sh		Script1_sample_FLAIR_smooth_features.sh
Script2_local_cortical_deformation.m		Script2_local_cortical_deformation.m
Script3_doughnut_method.m		Script3_doughnut_method.m
Script4_smooth_Doughnut_LCD.sh		Script4_smooth_Doughnut_LCD.sh
Script6_move_to_xhemi_flip.sh		Script6_move_to_xhemi_flip.sh
Script7_Z_Score_features_by_controls.m		Script7_Z_Score_features_by_controls.m
Script8_NeuralNetwork.m		Script8_NeuralNetwork.m
Script9_Cluster_NN_output.m		Script9_Cluster_NN_output.m
Scripts_5_intrasubject_normalisation.m		Scripts_5_intrasubject_normalisation.m
Supplementary_script_1_register_FLAIR_to_T1.sh		Supplementary_script_1_register_FLAIR_to_T1.sh
Supplementary_script_2_convert_label_to_mgh.m		Supplementary_script_2_convert_label_to_mgh.m

Uh oh!

Latest commit

History

Folders and files

Repository files navigation

FCDdetection

About

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages

Uh oh!

Latest commit

History

Folders and files

Repository files navigation

FCDdetection

About

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages

Uh oh!

Latest commit

History

Folders and files

Repository files navigation

FCDdetection

About

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages

Uh oh!

Latest commit

History

Folders and files

Repository files navigation

FCDdetection

About

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages

Uh oh!

Latest commit

History

Folders and files

Repository files navigation

FCDdetection

About

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages

Uh oh!

Latest commit

History

Folders and files

Repository files navigation